User Electricity Behavior Analysis Based on K-Means Plus Clustering Algorithm

Abstract

In order to help electricity companies to describe the behavior of customers accurately and guide the electricity department to adjust the power generation strategy effectively, a K-means plus clustering algorithm based on Python is proposed to classify the power consumption data in Taiyuan. By extracting the electricity data of company, the most suitable clustering number K is found. The K-means plus clustering algorithm classifies the data of electricity consumption and finally gets five different kinds of users. And then, the economic conditions of the users' households are analyzed. It is verified that the K-Means plus clustering algorithm is faster than K-means and the clustering result is more accurate.